Wrong number
Accounting for American public pensions is still flawed
Jun 30th 2011
from the print edition
THERE is crazy and then there is accounting for American public-sector pensions. A proposed set of reforms from the Government Accounting Standards Board (GASB), the standard-setter for America’s state and local governments, is a step in the right direction. But it will still leave an unsatisfactory system.
The best way to illustrate the insanity is to compare the public and private sectors. When a private company switches its pension scheme from a final-salary, or defined benefit (DB), basis to a defined-contribution (DC) basis, the intention is to save money. The company’s exposure to a DC scheme is limited to the amount that it puts in; in a DB scheme, it must make up any shortfall in investment returns.
A state that switches from DB to DC for new employees would find that its cash costs rise in the short term. New hires will make their contributions to the new DC scheme, not the old DB scheme, but the state still has to cough up for existing DB benefits. The current accounting rules force the state to recognise those costs more quickly once the DB scheme is closed to new members, dissuading some states from making the move. In Nevada a study by the Segal Group, a consultancy, found that employer contributions would have to rise by 10% of payroll, or $1.2 billion, over two years.
A switch from DB to DC will save money over the long term by capping the employer’s liability. But not that much. That is because many states have legal and constitutional provisions protecting existing workers’ pension rights. Such provisions safeguard not just the accrued pension rights of workers; they also bar states from altering their rights with regard to their future service. The same protections do not apply in the private sector.
The cast-iron nature of this pensions guarantee ought, you might think, to be reflected in pension accounting. But states are allowed a more generous accounting treatment than private-sector employers. They can discount their liabilities on the basis of the assumed rate of return (8% is standard) on the assets in their pension funds. Companies have to use the (lower) yield on an AA-rated corporate bond. So the present value of a state’s liabilities is lower than it would be if private-sector rules applied, even though the rights of public-sector workers are greater.
The GASB proposals do at least suggest that the unfunded portion of a state’s liability be discounted at an AA rate. But calculating this portion still involves the use of the expected return on assets.
The assumption that the entire pension fund can earn 8% is highly questionable. Given the current yields on cash and Treasury bonds, it requires a double-digit return from equities over the long term. That looks improbable. The return from equities comprises the dividend yield, plus dividend growth, plus or minus any change in the rating (the price-dividend ratio). Given that the starting dividend yield is 2%, you need to make some heroic assumptions to get to a double-digit return.
Whatever the excess return from equities, it is a reward for risk. You cannot simply assume this risk away, especially after the experience of the past ten years. After all, if states could count on a high return from equities, why stop at their pension-fund exposure? They should borrow money in the bond market and invest the proceeds in shares. They could use the profits to pay for future expenditure on schools and highways. Invest enough and perhaps state taxes could be abolished altogether. There is a chance, however, that citizens would consider this reckless.
In the past, high-return assumptions have allowed the states to get away with making lower contributions. But when the strategy goes wrong, the taxpayer foots the bill. A recent study by the Centre for Retirement Research at Boston College found that states have a funding ratio (the proportion of assets to liabilities) of 77% on the official measure, but just 51% if a risk-free rate is used.
This is a big hole to fill. A new paper* by Robert Novy-Marx of the University of Rochester and Joshua Rauh of the Kellogg School of Management in Chicago calculates that, to fully fund state and local pensions within 30 years, contributions will have to more than double. That translates into a tax increase of $1,398 per household per year, with five states (New Jersey, New York, Oregon, Wyoming and Ohio) requiring an increase of more than $2,000 a year. Taxpayers may be happy to make this contribution to the welfare of their fellow citizens. But they should at least be told about the size of the bill.
* “The Revenue Demands of Public Employee Pension Promises”, June 2011
0 comments:
Publicar un comentario